Project Summary Acute respiratory infections (ARIs) are caused by a number of bacterial, viral, and fungal pathogens and pathogen identification is needed to administer the correct treatment. However, due to the lagtime in standard biochemical assays and antimicrobial susceptibility testing, clinicians have come to depend on broad-spectrum, empirical treatment which contributes to both drug resistance and patient death. The goal of this project is to develop a breath test for rapid pathogen identification and treatment monitoring in ARI to facilitate more immediate, targeted treatment in the clinic. The current proposal is focused on bacterial pathogen identification with future goals to expand to viral and fungal pathogens. Substrate cleavage assays are currently used to query bacterial protease activity and can be used to rapidly and accurately classify bacteria down to the species level. Leveraging the protease-responsive nanosensor platform in the Bhatia lab, the goal of this proposal is to develop inhalable multiplexed nanosensors that release volatile reporters into the breath in response to infection- associated proteases in the lung. From there, we can generate breath ?fingerprints? for pathogens common in ARIs such as ventilator-associated pneumonia (VAP) (P. aeruginosa, S. aureus, K. pneumoniae, E. coli, S. pneumoniae, and H. influenzae). Furthermore, changes in proteolytic activity after the start of antimicrobial treatment can be used to generate a ?good response? and ?poor response? breath signature for more timely evaluation of drug efficacy. To this end, the specific aims of this project are the following: (1) establish a volatile- barcoding system for peptide substrates (2) build and validate an inhalable multiplexed system of protease nanosensors for pathogen identification and (3) investigate use of multiplexed protease nanosensors for monitoring response to antibiotic treatment. Aim 1 will be completed by identifying volatile reporter candidates that can be attached to peptide substrates without deleterious effects on cleavage kinetics, protease specificity, and breath signal. Once identified, volatile reporters will be isotope-labeled to create a panel of reporters with similar volatility differing only by mass. Peptides with orthogonal susceptibility to host and pathogen proteases will then be identified in Aim 2 by screening a peptide library against bacterial culture supernatants and bronchioalveolar lavage from infected mice. Peptides will then be barcoded using the VOC mass labels designed in Aim 1 and then formulated into inhalable nanosensors by attachment to a nanoparticle core. The resulting nanosensor panel will be delivered via intratracheal injection into mice infected with one of the six VAP pathogens. Machine learning will be used to generate a statistical classifier to identify pathogens based on reporter levels in breath and will be used to further identify reporter signatures for good and poor response to antibiotic treatment. Successful completion of these aims would result in a diagnostic platform that can potentially be expanded for rapid identification of an exhaustive list of respiratory pathogens, including viral and fungal pathogens.